Google’s Project Suncatcher Aims to Take AI to Space by 2027 to Scale Machine Learning The Bridge Chronicle
Tech

Google’s Project Suncatcher Aims to Take AI to Space by 2027 to Scale Machine Learning

Google’s Project Suncatcher plans to launch solar-powered AI chips into orbit by 2027, testing if machine learning can scale beyond Earth.

Manaswi Panchbhai

Google is taking artificial intelligence where no company has gone before —outer space. On November 4, 2025, the company announced Project Suncatcher, an ambitious plan to build solar-powered satellite constellations equipped with its Tensor Processing Units (TPUs) to run machine learning models beyond Earth’s atmosphere.

Join our WhatsApp Channel to Stay Updated!

The initial stage is set to commence in early 2027, with Google planning to deploy two prototype satellites in partnership with Planet Labs, a satellite imaging company based in the United States. This mission aims to evaluate if Google's TPUs can withstand the harsh conditions of low-Earth orbit and if optical links between satellites can facilitate distributed AI tasks.

A Leap Into the Future

Led by Travis Beals, Senior Director of Paradigms of Intelligence at Google, Project Suncatcher is described as “a moonshot grounded in data.” Beals says the early analysis shows “no fundamental barriers” preventing space-based AI compute from becoming reality.

“Our TPUs are headed to space,” said CEO Sundar Pichai. “Like any moonshot, it’s going to take years of testing and breakthroughs,” Pichai added. “But our 2027 launch with Planet will be a crucial step toward that vision.” If successful, Project Suncatcher could redefine not just how AI runs but where it lives. For Google, the next frontier of computation might truly be among the stars.

Why Take AI to Space?

At the heart of Project Suncatcher is a pursuit of limitless solar energy. In orbit, solar panels can generate up to eight times more power than on Earth and operate almost nonstop, cutting battery dependence.

Google envisions compact constellations of interconnected satellites forming what could evolve into data-center-scale compute infrastructure in space, reducing pressure on Earth’s energy resources while unlocking massive processing capacity for AI.

The Bigger Picture

With launch costs expected to fall below $200 per kilogram by the mid-2030s, Google believes space-based data centers could soon rival Earth-based ones as AI power demands soar.

The company’s approach also signals a growing trend. Amazon’s Jeff Bezos has forecasted space-based data centers within two decades, while startups like Starcloud, backed by Google and NVIDIA, are already experimenting in the same direction.

Shortage of Technical Equipment and Skilled Manpower Delays Expressway Gas Tanker Clearance by 32 Hours

Pune Bar Association Election Sees 50% Turnout, Management Draws Criticism

Mumbai–Pune Expressway Reopens After 32 Hours; Commuters Frustrated by Long Delays and Lost Work Hours

Pune Traffic: Dive Ghat to Remain Close For Three Hours Today For Blasting

Fourth Dimension Media Solutions Celebrates 15 Years of Media Excellence

SCROLL FOR NEXT